The term "boxcar" is very generic. There were many different builders and some RRs built some in their own shops. There were also many different sizes from 36 ft woodside to the humongous 80 ft high cube auto cars.
So to be very scale accurate, the person has to decide what RR and then research what that particular RR did to get the specific size boxcar in question.
A nice paint job is the last part of the evaluation of a given model. The first should be does it represent the correct prototype, ACF or Pullman Standard or whoever, and size. Then the details should be considered based on era. Does it have roof walks, and if so are they wood planks or steel mesh? Does it have the right kind of doors of correct width (6 - 8 - 10 ft) Does it have the right ends, lots of differences here. then does it have the right roof panels? the trucks should also be considered but they can be changed. Another issue is the level of fine details such as grabs and stirrups and brake rigging. Now here, more is obviously closer to scale, but depending on use cast on may make more sense to avoid breakage. Personally I have been moving away from cast on details. In O scale individual grabs and rigging are obvious to my eye very plastic slabs.
But the most important things to consider are the fidelity to scale size and proportions of the body.
